In this paper, we report our extension of the use of feature selection based on mutual information and feature fusion to improve gender classification of face images. We compare the results of fusing three groups of features, three spatial scales, and four different mutual information measures to select features. We also showed improved results by fusion of LBP features with different radii and spatial scales, and the selection of features using mutual information. As measures of mutual information we use minimum redundancy and maximal relevance (mRMR), normalized mutual information feature selection (NMIFS), conditional mutual information feature selection (CMIFS), and conditional mutual information maximization (CMIM). We tested the results on four databases: FERET and UND, under controlled conditions, the LFW database under unconstrained scenarios, and AR for occlusions. It is shown that selection of features together with fusion of LBP features significantly improved gender classification accuracy compared to previously published results. We also show a significant reduction in processing time because of the feature selection, which makes real-time applications of gender classification feasible.

Previous researchers have explored various approaches for predicting the gender of a person based on the features of the iris texture. This paper is the first to predict gender directly from the same binary iris code that could be used for recognition. We found that the information for gender prediction is distributed across the iris, rather than localized in particular concentric bands. We also found that using selected features representing a subset of the iris region achieves better accuracy than using features representing the whole iris region. We used the measures of mutual information to guide the selection of bits from the iris code to use as features in gender prediction. Using this approach, with a person-disjoint training and testing evaluation, we were able to achieve 89% correct gender prediction using the fusion of the best features of iris code from the left and right eyes.

An accurate and reliable image-based quantification system for blueberries may be useful for the automation of harvest management. It may also serve as the basis for controlling robotic harvesting systems. Quantification of blueberries from images is a challenging task due to occlusions, differences in size, illumination conditions and the irregular amount of blueberries that can be present in an image. This paper proposes the quantification per image and per batch of blueberries in the wild, using high definition images captured using a mobile device. In order to quantify the number of berries per image, a network based on Mask R-CNN for object detection and instance segmentation was proposed. Several backbones such as ResNet101, ResNet50 and MobileNetV1 were tested. The performance of the algorithm was evaluated using the Intersection over Union Error (IoU) and the competitive mean Average Precision (mAP) per images and per batch. The best detection result was obtained with the ResNet50 backbone achieving a mIoU score of 0.595 and mAP scores of 0.759 and 0.724 respectively (for IoU thresholds 0.5 and 0.7). For instance segmentation, the best results obtained were 0.726 for the mIoU metric and 0.909 and 0.774 for the mAP metric using thresholds of 0.5 and 0.7 respectively.

Currently, it is ever more common to access online services for activities which formerly required physical attendance. From banking operations to visa applications, a significant number of processes have been digitised, especially since the advent of the COVID-19 pandemic, requiring remote biometric authentication of the user. On the downside, some subjects intend to interfere with the normal operation of remote systems for personal profit by using fake identity documents, such as passports and ID cards. Deep learning solutions to detect such frauds have been presented in the literature. However, due to privacy concerns and the sensitive nature of personal identity documents, developing a dataset with the necessary number of examples for training deep neural networks is challenging. This work explores three methods for synthetically generating ID card images to increase the amount of data while training fraud-detection networks. These methods include computer vision algorithms and Generative Adversarial Networks. Our results indicate that databases can be supplemented with synthetic images without any loss in performance for the print/scan Presentation Attack Instrument Species (PAIS) and a loss in performance of 1% for the screen capture PAIS.

Quality assessment algorithms can be used to estimate the utility of a biometric sample for the purpose of biometric recognition. “Error versus Discard Characteristic” (EDC) plots, and “partial Area Under Curve” (pAUC) values of curves therein, are generally used by researchers to evaluate the predictive performance of such quality assessment algorithms. An EDC curve depends on an error type such as the “False Non Match Rate” (FNMR), a quality assessment algorithm, a biometric recognition system, a set of comparisons each corresponding to a biometric sample pair, and a comparison score threshold corresponding to a starting error. To compute an EDC curve, comparisons are progressively discarded based on the associated samples’ lowest quality scores, and the error is computed for the remaining comparisons. Additionally, a discard fraction limit or range must be selected to compute pAUC values, which can then be used to quantitatively rank quality assessment algorithms. This paper discusses and analyses various details for this kind of quality assessment algorithm evaluation, including general EDC properties, interpretability improvements for pAUC values based on a hard lower error limit and a soft upper error limit, the use of relative instead of discrete rankings, stepwise vs. linear curve interpolation, and normalisation of quality scores to a 0, 100 integer range. We also analyse the stability of quantitative quality assessment algorithm rankings based on pAUC values across varying pAUC discard fraction limits and starting errors, concluding that higher pAUC discard fraction limits should be preferred. The analyses are conducted both with synthetic data and with real face image and fingerprint quality assessment data, with a focus on general modality-independent conclusions for EDC evaluations. Various EDC alternatives are discussed as well. Open source evaluation software is provided at https://github. com/dasec/quality-assessment-evaluation.

Iris pattern recognition has significantly improved the biometric authentication field due to its high stability and uniqueness. Such physical characteristics have played an essential role in security applications and other related areas. However, presentation attacks, also known as spoofing techniques, can bypass biometric authentication systems using artefacts such as printed images, artificial eyes, textured contact lenses, etc. Many liveness detection methods that improve the robustness of these systems have been proposed. The first International Iris Liveness Detection competition, where the effectiveness of liveness detection methods is evaluated, was first launched in 2013, and its latest iteration was held in 2020. In this paper, we present the approach that won the LivDet-Iris 2020 competition using two-class scenarios (bona fide iris images vs. presentation attack iris images). Additionally, we propose new three-class and four-class scenarios that complement the competition results. These methods use a serial architecture based on a MobileNetV2 modification, trained from scratch to classify bona fide iris images versus presentation attack images. The bona fide class consists of live iris images, whereas the attack presentation instrument classes consist of cadaver, printed, and contact lenses images, for a total of four species. All the images were pre-processed and weighted per class to present a fair evaluation. This approach is primarily focused on detecting the bona fide class over improving the detection of presentation attack instruments. For the two, three, and four classes scenarios BPCER ₁₀ values of 0.99%, 0.16%, and 0.83% were obtained respectively, whereas for the BPCER ₂₀ values of 3.09%, 0.16%, and 3.77% were obtained, with the best model overall being the proposed 3-class serial model. This work reaches competitive results according to the reported results in the LivDet-Iris 2020 competition.